Conventional computed tomography (CT) scanners and digital radiography systems use large numbers of X-ray detectors, on the order of several hundred to several thousand, in which each X-ray detector includes a scintillator to convert X-rays into light and a photocell, or photodiode array, to convert light into an electrical signal.
Certain photodiodes have two-electrode radiation-sensitive junctions formed in semiconductor material. Light, which illuminates the junction, creates charge carriers (via mobile or “free” electrons and holes). Doping a semiconductor with a small amount of impurity greatly increases the number of charge carriers within the semiconductor. When a doped semiconductor has excess (majority) holes, it is called p-type and when it contains excess (majority) free electrons, it is known as n-type. The holes in a p-dopes semiconductor are majority carriers while the electrons are minority carriers. In the case of n-type doping, the electrons are the majority carriers and the holes are the minority carriers. The junctions which form where n-type and p-type semiconductors join together are called P-N junctions. At the P-N junction, there forms a region called the depletion zone, which blocks current conduction from the n-type region to the p-type region, but allows current to conduct from the p-type region to the n-type region. The depletion region is void of all majority carriers and is a non-conducting layer. In other words, the recombination of holes and electrons at the P-N junction causes the region to become depleted of mobile charge.
The depletion region forms across the P-N junction when the junction is in thermal equilibrium, i.e. things are in a steady state. Electrons and holes will diffuse into regions with lower concentrations of electrons and holes. N-type semiconductors have an excess of free electrons while P-type semiconductors have an excess of holes. Therefore, when a P-N junction is formed, electrons will diffuse into the P side and holes will diffuse into the N side. When a hole and an electron come into contact, however, they eliminate each other through recombination. This bares the donor atoms adjacent to the depletion region, which are now charged ions. The ions are positive on the N side and negative on the P side, creating an electric field that counteracts the continued diffusion of charge carriers. When the electric field is sufficient to repel incoming holes and electrons, the depletion region reaches its equilibrium width.
Under reverse bias (P negative with respect to N) this potential is increased, further widening the depletion zone. Forward bias (P positive with respect to N) narrows the zone and eventually reduces it to nothing, making the junction conductive and allowing free flow of charge carriers. Thus, it is possible to manipulate the non-conductive layer to allow a flow of electricity in one direction but not the other (opposite) direction. When the P-N junction is forward-biased, electric charge flows freely due to reduced resistance of the P-N junction. When the P-N junction is reverse biased, however, the junction barrier (and therefore resistance) becomes greater and charge flow is minimal.
Essentially active solid-state semiconductor devices, and in particular, silicon photodiodes, are among the most popular photodetectors having a sufficiently high performance over a large wavelength range with ease of use. Silicon photodiodes are sensitive to light in the wide spectral range, extending from deep ultraviolet through visible to near infrared, which is approximately 200 nm to 1100 nm. Silicon photodiodes, by using their ability to detect the presence or absence of minute light intensities, facilitate the extremely precise measurement of these minute light intensities upon appropriate calibration. For example, appropriately calibrated silicon photodiodes detect and measure light intensities varying over a wide range, from very minute light intensities of below 10−13 watts/cm2 to high intensities above 10−3 watts/cm2.
Silicon photodiodes can be employed in an assortment of applications including, but not limited to, spectroscopy, distance and speed measurement, laser ranging, laser guided missiles, laser alignment and control systems, optical free air communication, optical radar, radiation detection, optical position encoding, film processing, flame monitoring, scintillator read out, environmental applications such as spectral monitoring of earth ozone layer and pollution monitoring, low light-level imaging, such as night photography, nuclear medical imaging, photon medical imaging, and multi-slice computer tomography (CT) imaging, security screening and threat detection, thin photochip applications, and a wide range of computing applications.
Typically, photodiode arrays employ a scintillator material for absorbing high energy (ionizing) electromagnetic or charged particle radiation, which, in response, fluoresces photons at a characteristic wavelength. Scintillators are defined by their light output (number of emitted photons per unit absorbed energy) short fluorescence decay times, and optical transparency at wavelengths of their own specific emission energy. The lower the decay time of a scintillator, that is, the shorter the duration of its flashes of fluorescence are, the less so-called “dead time” the detector will have and the more ionizing events per unit of time it will be able to detect. Scintillators are used to detect electromagnetic waves or particles in many security and detection systems, including CT, X-ray, and gamma ray. There, a scintillator converts the energy to light of a wavelength which can be detected by photomultiplier tubes (PMTS) or P-N junction photodiodes.
Photodiodes are typically characterized by certain parameters, such as, among others, electrical characteristics, optical characteristics, current characteristics, voltage characteristics, and noise. Electrical characteristics predominantly comprise shunt resistance, series resistance, junction capacitance, rise or fall time and/or frequency response. Optical characteristics comprise responsivity, quantum efficiency, non-uniformity, and/or non-linearity. Photodiode noise may comprise, among others, thermal noise, quantum, photon or shot noise, and/or flicker noise.
In an effort to increase the signal to noise ratio and enhance the contrast of the signal, it is desirable to increase the light-induced current of photodiodes. Thus, photodiode sensitivity is enhanced while the overall quality of the photodiode is improved. Photodiode sensitivity is crucial in low-level light applications and is typically quantified by a parameter referred to as noise equivalent power (NEP), which is defined as the optical power that produces a signal-to-noise ratio of one at the detector output. NEP is usually specified at a given wavelength over a frequency bandwidth.
Photodiodes absorb photons or charged particles, facilitating detection of incident light or optical power and generating current proportional to the incident light, thus converting the incident light to electrical power. Light-induced current of the photodiode corresponds to the signal while “dark” or “leakage” current represents noise. “Dark” current is that current that is not induced by light, or that is present in the absence of light. Photodiodes process signals by using the magnitude of the signal-to-noise ratio.
Leakage current is a major source of signal offset and noise in current photodiode array applications. Leakage current flows through the photodiode when it is in a “dark” state, or in the absence of light at a given reverse bias voltage applied across the junction. Leakage current is specified at a particular value of reverse applied voltage. Leakage current is temperature dependent; thus, an increase in temperature and reverse bias results in an increase in leakage or dark current. A general rule is that the dark current will approximately double for every 10° C. increase in ambient temperature. It should be noted, however, that specific diode types can vary considerably from this relationship. For example, it is possible that leakage or dark current will approximately double for every 6° C. increase in temperature.
Various approaches have been used in the prior art to reduce, eliminate or control leakage current. For example, U.S. Pat. No. 4,904,861, assigned to Agilent Technologies, Inc., discloses “[an] optical encoder comprising: a plurality of active photodiodes in an array on a semiconductor chip; a code member having alternating areas for alternately illuminating and not illuminating the active photodiodes in response to movement of the code member; means connected to the active photodiodes for measuring current from the active photodiodes; and sufficient inactive photodiode area on the semiconductor chip at each end of the array of active photodiodes to make the leakage current to each end active photodiode of the array substantially equal to the leakage current to an active photodiode remote from an end of the array”. Similarly, U.S. Pat. No. 4,998,013, also assigned to Agilent Technologies, Inc. discloses “means for shielding a photodiode from leakage current comprising: at least one active photodiode on a semiconductor chip; means for measuring current from the active photodiode; a shielding area having a photodiode junction substantially surrounding the active photodiode; and means for biasing the shielding area photodiode junction with either zero bias or reverse bias.”
U.S. Pat. No. 6,670,258, assigned to Digirad Corporation, discloses “[a] method of fabricating a low-leakage current photodiode array comprising: defining frontside structures for a photodiode on a front side of a substrate; forming a heavily-doped gettering layer on a back surface of the substrate; carrying out a gettering process on the substrate to transport undesired components from the substrate to said gettering layer, and to form another layer in addition to said gettering layer, which is a heavily-doped, conductive, crystalline layer within the substrate; after said gettering process, removing the entire gettering layer; and after said removing, thinning the heavily-doped, conductive, crystalline layer within the substrate to create a native optically transparent, conductive bias electrode layer”. Similarly, U.S. Pat. No. 6,734,416, also assigned to Digirad Corporation, discloses “[a] low-leakage current photodiode array comprising: a substrate having a front side and a back side; a plurality of gate regions formed near the front side of the substrate; a backside layer formed within the substrate, near the back side of the substrate, the backside layer having a thickness of approximately 0.25 to 1.0 micrometers and having a sheet resistivity of approximately 50 to 1000 Ohm per square.”
U.S. Pat. No. 6,569,700, assigned to United Microelectronics Corporation in Taiwan, discloses “[a] method of reducing leakage current of a photodiode on a semiconductor wafer, the surface of the semiconductor wafer comprising a p-type substrate, a photosensing area for forming a photosensor of the photodiode, and a shallow trench positioned in the substrate surrounding the photosensing area, the method comprising: forming a doped polysilicon layer containing p-type dopants in the shallow trench; using a thermal process to cause the p-type dopants in the doped polysilicon layer to diffuse into portions of the p-type substrate that surround a bottom of the shallow trench and walls of the shallow trench; removing the doped polysilicon layer; filling an insulator into the shallow trench to form a shallow trench isolation (STI) structure; performing a first ion implantation process to form a first n-type doped region in the photosensing area; and performing a second ion implantation process to form a second n-type doped region in the photosensing area.”
Also, U.S. Pat. No. 6,504,158, assigned to General Electric Company, discloses “a method of reducing leakage current in an imaging apparatus, including: providing a substrate with at least one radiation-sensitive imaging region therein; forming a guard region in the substrate at or immediately adjacent a cut edge of the substrate to reduce leakage current reaching the at least one radiation-sensitive imaging region from the cut edge when the imaging apparatus is in use; and electrically reverse biasing the at least one radiation-sensitive imaging region and the guard region relative to the substrate.”
In certain applications, it is desirable to produce optical detectors having small lateral dimensions and spaced closely together. For example in certain medical applications, it is desirable to increase the optical resolution of a detector array in order to permit for improved image scans, such as computed tomography (CT) scans. However, at conventional doping levels utilized for diode arrays of this type, the diffusion length of minority carriers generated by photon interaction in the semiconductor is in the range of at least many tens of microns, and such minority carriers have the potential to affect signals at diodes away from the region at which the minority carriers were generated.
Thus, an additional disadvantage with conventional photodiode arrays is the amount and extent of crosstalk that occurs between adjacent detector structures, primarily as a result of minority carrier leakage current between diodes. The problem of crosstalk between diodes becomes even more acute as the size of the detector arrays, the size of individual detectors, the spatial resolution, and spacing of the diodes is reduced.
Various approaches have been used to minimize such crosstalk including, but not limited to, providing inactive photodiodes to balance the leakage current, as described in U.S. Pat. Nos. 4,904,861 and 4,998,013 to Epstein et al., the utilization of suction diodes for the removal of the slow diffusion currents to reduce the settling time of detectors to acceptable levels, as described in U.S. Pat. No. 5,408,122, and providing a gradient in doping density in the epitaxial layer, as described in U.S. Pat. No. 5,430,321 to Effelsberg.
In addition to leakage current and effects of crosstalk, noise is often a limiting factor for the performance of any device or system. In almost every area of measurement, the limit to the detectability of signals is set by noise, or unwanted signals that obscure the desired signal. As described above, the NEP is used to quantify detector noise. Noise issues generally have an important effect on device or system cost. Conventional photodiodes are particularly sensitive to noise issues. Like other types of light sensors, the lower limits of light detection for photodiodes are determined by the noise characteristics of the device.
As described above, the typical noise components in photodiodes include thermal noise; quantum or shot noise; and flicker noise. These noise components collectively contribute to the total noise in the photodiode. Thermal noise, or Johnson noise, is inversely related to the value of the shunt resistance of photodiode and tends to be the dominant noise component when the diode is operated under zero applied reverse bias conditions. Shot noise is dependent upon the leakage or dark current of photodiode and is generated by random fluctuations of current flowing through the device, which may be either dark current or photocurrent. Shot noise tends to dominate when the photodiode is used in photoconductive mode where an external reverse bias is applied across the device. As an example, detector noise generated by a planar diffused photodiode operating in the reverse bias mode is a combination of both shot noise and thermal noise. Flicker noise, unlike thermal or shot noise, bears an inverse relationship to spectral density. Flicker noise may dominate when the bandwidth of interest contains frequencies less than 1 kHz.
Secondary issues also contribute to dark noise and other noise sources that impact photodiode sensitivity. These include primarily determination and/or selection of apt active area specifications (geometry and dimensions), response speed, quantum efficiency at the wavelength of interest, response linearity, and spatial uniformity of response, among others.
In CT applications, such as those employed for baggage screening, it is desirable to have high density photodiode arrays with low dark current, low capacitance, high signal-to-noise ratio, high speed and low crosstalk.
As mentioned above, however, there are numerous problems with conventional photodiodes that attempt to achieve these competing and often conflicting characteristics. For example, in order to achieve low capacitance the photodiode can be fabricated on a high resistivity (on the order of 4000-6000 Ωcm) silicon material. Using a high resistivity material, however, causes the device to have high dark current.
FIG. 1 is a cross-sectional view of a conventional, prior art fishbone photodiode device 100. The photodiode array 100 comprises substrate wafer 105, which is a thick bulk wafer having an active area thickness on the order of 275-400 μm. As shown in FIG. 1 on the conventional fishbone photodiode device some photogenerated holes 110 move randomly in various directions, such as paths 115, in the thick active volume 106 of the bulk starting material wafer. Since minority carrier lifetime is limited, many of these photogenerated holes are lost due to recombination of holes and electrons in the bulk material, which causes a reduction in charge collection efficiency or responsivity of the photodiode.
In order to improve charge collection efficiency in this prior art fishbone photodiode device, the p+ diffused bones need to be placed relatively close to each other. This is disadvantageous, however, because a relatively large number of p+ fishbones is needed when placing the p+ diffused bones closer together, resulting in high junction capacitance. Typically, charge carriers that are photo-generated further from the P-N junction can diffuse toward the p+ diffused “bones” and be collected by the depletion region.
In addition, the fishbone photodiode device in the prior art as shown in FIG. 1 is disadvantageous because the high volume, thick active layer 105 that is used to fabricate the photodiode results in high dark current, since dark current is proportional to the overall volume of the active layer of the device.
In addition, the conventional photodiode array described above with respect to FIG. 1 is disadvantageous in that the photodiode tends to degrade in shunt resistance since the P-N junction is passivated by relatively thin antireflective layers, such as silicon oxide on the order of 150 Å and silicon nitride on the order of 425 Å.
The result is a fishbone photodiode having high noise characteristics, and thus, a poor signal to noise ratio.
What is needed is a photodiode array that can be fabricated on a thin active layer. In particular, what is needed is a photodiode array that can be fabricated on a thin active layer such as thin epi or thin direct-bonded layer, for fast rise time and better charge collection efficiency.
What is also needed is a photodiode array having reduced junction capacitance and reduced dark current, thus improving the signal to noise ratio of the photodiode array.
What is also needed is a photodiode array having reduced junction capacitance and reduced dark current, thus improving the signal to noise ratio of the photodiode array without sacrificing performance characteristics, such as quantum efficiency.
In addition, what is needed is economically, technically, and operationally feasible methods, apparatuses, and systems for manufacturing photodiode arrays on a thin active layer with reduced junction capacitance and reduced dark current effects.
In addition, what is needed is economically, technically, and operationally feasible methods, apparatuses, and systems for manufacturing photodiode arrays that can be used in computed tomography (CT) scanner applications that improve upon overall performance characteristics of the photodiode array and individual diode elements.